QUOTE (jmvalin @ Apr 19 2005, 06:26 AM)
QUOTE (sevana @ Apr 19 2005, 12:06 AM)
We are working in the field of audio and speech technologies. We have developed a software that can automatically compare two audio codecs. We would be happy if you consider this software and come back to us with your feedback. Thank you very much in advance! Your opinion is very important for our research!
Just looking at the results, something's really odd. It would seem to indicate that G.723.1 6.3 kbps is better than G.729A and that G.723.1 5.3 kbps is about the same quality. Since these codecs are a bit similar (G.723.1 being a bit older), I'd be surprised if that was really the case. Seeing G.728 score even below that is very strange too. I'm not sure where these MOS come from, but they're really suspicious. Note that you can't compare MOS values that are part of two different tests (MOS aren't absolute).
Hi,
Thanks a lot for your feedback!
Our program is using adapted IEC 268-16, so our tests are made using artificial signal, not real speech. Furthermore, same codecs may give different MOS estimations for different languages as the speech consistency is different.
There is a yet another aspect... what we are trying to do is to have a more or less objective and automatic codec quality analyzer, so that codec developers will not need to spend lots of time (and money) comparing two codecs.
If I didn't understand your point correctly, please let me know

Thank you all for your time and feedback!
Regards,
Sevana OY
Finland